Emerald AI Study Shows Data Centers Can Reduce Energy Consumption by 25%
A study by Emerald AI reveals that AI data centers can achieve a 25% reduction in energy consumption during peak periods without compromising performance. As energy demand for these data centers in Europe is projected to rise significantly by 2030, the research underscores the importance of integrated solutions among network operators, data centers, and end users. Key findings will be presented at Datacloud Energy Europe in Brussels on March 25-26, 2026.
Theia Market Signal Identification - AI Assisted
A study by Emerald AI researchers indicates that AI data centers can improve energy efficiency without sacrificing performance. The study predicts that total load demand for AI data centers in Europe will rise from 10 gigawatts in 2024 to 35 gigawatts by 2030, with annual energy consumption increasing from 62 terawatt-hours (TWh) to over 150 TWh, representing 5% of Europe's total electricity consumption.
In China, energy consumption is projected to reach 600 TWh by 2030. The research, conducted in collaboration with SRP, Nvidia, Oracle Corporation, and EPRI, demonstrated that a software solution could modulate processing tasks to reduce energy consumption by 25% during peak stress periods without significant performance impacts.
This approach emphasizes the need for integrated solutions involving network operators, data centers, and end users. The findings will be discussed at Datacloud Energy Europe in Brussels on March 25-26, 2026.
Comments